China Securities Journal on Chips: Continuous breakthroughs, domestic AI chips seize market window, industry chain accelerates technological iteration

```

From Alibaba announcing an investment of 380 billion yuan over the next three years, to Baidu Kunlun Chip winning a multi-billion yuan order from China Mobile... Amidst the global surge in demand for AI computing power, the Chinese domestic AI chip industry is ushering in a critical strategic opportunity.

On September 16, "China Securities Journal" published two in-depth articles analyzing the latest developments in China’s chip industry. The articles believe that domestic chip manufacturers and the optical communication industry chain are seizing this key "window period" to accelerate technology iteration and market deployment.

Domestic AI Chips Seize Market Window

According to "China Securities Journal" citing industry insiders, the strong market demand for AI chips is "driving domestic chip manufacturers and Internet giants to fully seize the market by taking advantage of the unstable supply of Nvidia chips during this window period."

The article points out that a Bernstein research report released in July estimates that domestic AI chip demand will reach 39.5 billion USD by 2025, and the market localization rate will increase from 17% in 2023 to 55% in 2027.

Behind this trend are the large investments of domestic Internet giants and the accelerated commercialization of domestic chip manufacturers.

Big Companies Increasing Investment: The article mentions Alibaba has announced "an investment of 380 billion yuan over the next three years to build cloud and AI hardware infrastructure." CEO Wu Yongming revealed that in the past four quarters, the company has invested over 100 billion yuan in AI infrastructure and product R&D. Meanwhile, Baidu’s Kunlun Chip also announced winning a multi-billion yuan order from China Mobile, becoming “an important case of major internet companies’ self-developed chips winning large external orders.”
Impressive Performance: Demand is directly translating into results for domestic chip manufacturers. According to the article, Cambricon’s revenue in the first half of the year shot up by 4347.82% year-on-year, mainly contributed by its cloud product line; Hygon’s revenue in the first half rose 45.21%, and the company states that “demand for domestic high-end chips continues to increase.” VeriSilicon also announced that as of the end of Q2, orders in hand reached 3.025 billion yuan, with about 64% of new orders related to AI computing power.

Industry insiders said that companies such as Cambricon and Hygon “already have strong capabilities in AI inference chips.” The urgent task is to quickly ramp up production capacity. To this end, companies such as Huahong, SMIC, and Cambricon are accelerating expansion through mergers and acquisitions and targeted financing.

Computing Power Demand Drives Accelerated Industry Iteration

The explosive growth of AI chips is also placing higher demands on the upstream industry chain. The article notes, "the rapid growth of AI computing demand is driving rapid iteration of core components and packaging technology," with optical modules becoming the focus.

The article states that the development of the AI industry will "directly drive a surge in demand for optical modules, with 1.6T modules becoming the focus of competition." At a recent Optical Expo, manufacturers such as Luxshare-Tech and Accelink showcased cutting-edge solutions including co-packaged optics (CPO) and 1.6T optical modules.

Hui Jun, fund manager at E Fund’s equity investment research department, said, “With the dual drivers of AI computing clusters and supercomputing center construction, the global optical module market will experience a structural boom in 2025, and the market size is expected to reach 12.1 billion USD.” He predicts that CPO "will accelerate penetration and is expected to exceed 20% market share in AI data centers by 2026, becoming the mainstream solution for next-generation optical interconnects."

Strong demand has already been reflected in the financial reports of related listed companies. Innolight’s net profit in the first half of the year increased by 355.68% year-on-year; Accelink’s net profit rose 69.40%, and the company stated, “Demand for 800G and 1.6T products is growing rapidly.”

Challenges and Bottlenecks: Capacity and Ecosystem Building

Despite promising prospects, the articles also clearly point out the industry’s bottlenecks.

First is capacity and delivery. Guosheng Securities said that the industry challenge has shifted from “where is the demand” to “how to deliver on time.” Accelink also admitted during investor relations that “the supply of some key raw materials is tight.” To address these challenges, companies like Innolight and Tianfu Communications are accelerating the capacity expansion of their factories in Thailand.

Secondly, and more critically, is the software ecosystem. The article cites Professor Zhai Jidong from Tsinghua University’s Computer Science Department, saying that the domestic level of computing hardware is already close to or even surpasses competing Nvidia chips, but there is still room for improvement in the software ecosystem. He emphasized, “To perfect the software ecosystem, you need everything — from the system schedulers and memory management at the bottom, to programming languages, AI compilers, operator libraries in the middle, and programming frameworks at the top — all are essential.”

Alibaba Group CEO Wu Yongming also mentioned that in the future, the cloud computing market will become more concentrated and developers will tend to choose vendors “with a comprehensive technology and product portfolio.” A research report by Shanghai Securities points out that a sound software platform is key to harnessing hardware performance and lowering user barriers. In conclusion, the article notes that leading domestic enterprises are continuing to improve the AI chip ecosystem in order to further enhance market competitiveness.

Risk Reminder and DisclaimerThe market carries risks; investment should be cautious. This article does not constitute individual investment advice, nor does it take into account any specific user's investment objectives, financial situation, or needs. Users should consider whether any opinions, views, or conclusions in this article are appropriate to their particular circumstances. Investing based on this article is at your own risk. ```